Cultures And Organizations Software Of The Mind

Understanding Organizational Culture Before You Try to Change It

Most people who walk into a company and declare "we need better culture" have never actually understood what culture is. They see slow decision-making or awkward meetings and assume the problem is vague vibes. It isn't. Culture is a real system. It's the accumulated set of assumptions a group has solved over time, and it operates below the level of most employees' conscious awareness. When you try to fix something called "culture," you are really trying to modify a deep cognitive operating system that took years to install. Edgar Schein's model breaks culture down into three layers. The first and most obvious layer is artifacts. These are the things you can observe immediately: the dress code, the layout of the office, how people talk to each other in meetings, the documents on the wall, the way decisions get made. Artifacts are easy to see but nearly impossible to interpret correctly without understanding what's underneath them. A flat organization with ping pong tables and open floor plans might look like it has a relaxed culture, or it might just be a company spending money oncool perks while the actual power structure remains deeply hierarchical and unfriendly.

Cultures And Organizations Software Of The Mind

The second layer consists of espoused strategies, goals, and values. This is what people say the company stands for. You'll find it in mission statements, onboarding presentations, and leadership speeches. The problem with this layer is that it is often pure theater. Companies routinely claim to value "transparency" and "work-life balance" while systematically punishing people who speak up or leave at five. Espoused values tell you what the organization wants to believe about itself. They rarely tell you how it actually operates. The third and most important layer is the basic underlying assumptions. These are unconscious, taken-for-granted beliefs, perceptions, thoughts, and feelings. They are the real software running the organization. Things like "people are inherently lazy and need to be monitored," or "conflict should be avoided at all costs to preserve harmony," or "innovation requires protecting ideas until they are perfect before sharing them." These assumptions are invisible to most people in the organization because they formed so gradually and were reinforced so many times that no one questions them anymore. They dictate behavior more powerfully than any mission statement ever could. I spent about three years working with engineering teams trying to implement agile practices across several different organizations, and the patterns kept repeating in ways that had nothing to do with process. We would bring in Scrum training, set up boards, run retrospectives, and within six months everyone would go back to doing exactly what they had always done. The surface-level artifacts of agile would remain — standup meetings, story points, sprint planning — but underneath, the assumptions never shifted. In one specific case, a mid-size software company had a deeply held assumption that the senior architect's word was final and that questioning technical direction was a sign of weakness or disloyalty. We tried retrospectives, we tried anonymous feedback forms, we tried rotating facilitators. Nothing moved the needle. What eventually worked was a deliberate, uncomfortable intervention where leadership publicly admitted a major technical mistake they had made by not listening to junior engineers, and then tied their own compensation partially to measurable improvements in cross-team knowledge sharing. It took eight months of consistent reinforcement before the underlying assumption started to shift. Even now, two years later, it only shifts when someone actively maintains the pressure. If leadership goes quiet about it for six months, the old patterns resurface.

How to Work With This Model in Practice

The most useful thing about Schein's framework is that it gives you a diagnostic map. Instead of feeling confused by organizational dysfunction, you can ask systematic questions at each layer. When something feels wrong, start by describing the artifacts you observe. Then identify the espoused values that seem to contradict those artifacts. Then work backward to figure out what underlying assumptions must be true for those contradictions to persist. That gap between the espoused and the actual is where the real work lives. Here is a practical method for conducting a culture assessment that does not require a consulting budget. First, spend two weeks simply observing. Sit in meetings. Read internal communication. Note what gets rewarded and what gets punished. Don't judge anything yet. Second, interview ten to fifteen people across different levels and departments using the same five questions: What do we do here that would surprise someone from another company? What is not allowed to be said? How are promotions actually decided, not how the handbook says they are decided? What happens when someone makes a mistake? What do people really care about? Third, look for patterns in the answers that contradict the official values. Those patterns point directly to the underlying assumptions. Fourth, prioritize which assumptions are enabling the outcomes you want and which are blocking them. Most culture change efforts fail because they try to change everything at once. Pick one or two assumptions that are creating the most friction and focus exclusively on those. Changing an underlying assumption is not a communication problem. It is a behavioral conditioning problem. People hold an assumption because it has been reinforced as useful over hundreds of repeated experiences. To change it, they need new repeated experiences that contradict the old assumption and provide a better outcome. This means you have to design interventions that create those new experiences deliberately and consistently. It means changing reward structures, creating new rituals, removing old ones, and modeling new behaviors at the leadership level with zero exceptions. It usually takes eighteen to twenty-four months of sustained effort to see a genuine shift in a deeply held assumption. Anything shorter than that is just training people to say different words.

Common Pitfalls That Make This Worse

The single most destructive thing an organization can do is announce a cultural initiative without changing the underlying incentive structure. If you tell people "we value collaboration" but your performance reviews and bonuses are still based entirely on individual output, nobody will collaborate. They will collaborate when it is convenient and safe, but the moment there is any risk to their personal metrics, they will retreat to individual work. I have seen this play out identically in companies of fifty people and companies of fifty thousand. The pattern never changes because the mechanism is the same. Espoused values that are not backed by structural incentives are just noise. They actually make things worse because they create cynicism. People notice the gap between what is said and what is rewarded, and they stop believing anything the organization claims to value. Another trap is treating culture as something that can be designed from the top down. You can design artifacts. You can draft values. You can install processes. But you cannot directly design underlying assumptions because those emerge from collective experience over time. The best you can do is create conditions where certain assumptions become more likely to form and others become less likely. That is a subtle but critical distinction. Leading from the top works for changing policy. It does not work for changing deeply held assumptions without extensive time and repeated proof that the old assumptions are no longer useful. There is also a significant limitation to Schein's model that nobody talks about enough. It assumes that culture is relatively stable and coherent within an organization. In practice, many companies have multiple competing subcultures that are in active tension. A sales organization and an engineering organization in the same company can operate under completely different basic assumptions about risk, time horizons, and what counts as evidence. Treating the company as having one culture is a category error. You will miss the friction points that actually matter if you look for a single unified cultural profile. In one engagement, we spent four months trying to understand why a product company could not ship anything fast enough. The root cause was not a vague "culture problem." It was a direct collision between the engineering team's assumption that shipping incomplete work was unacceptable and the marketing team's assumption that missing a coordinated launch window was catastrophic. Both assumptions made perfect sense within their own subcultures. Resolving it required negotiating the boundaries between subcultures, not pretending a single culture existed.

When This Approach Does Not Apply

Schein's model is not useful in every situation. If you are trying to solve a tactical problem like low productivity, poor code quality, or high customer churn, start with those operational issues before you assume culture is the root cause. Culture is often a secondary effect of bad processes, unclear requirements, inadequate tooling, or misaligned incentives. Digging into culture first can become a form of organizational procrastination. It feels productive to talk about culture because it is abstract and nobody gets blamed directly. Fixing real operational problems is harder and requires specific technical knowledge. The model also breaks down in very small organizations, typically under twenty people, where culture is essentially identical to the founder's personal assumptions. There is no gap between espoused values and underlying assumptions because the founder says what they mean and means what they say. In those contexts, looking for cultural layers is academic exercise rather than practical diagnosis. You can just talk to the founder. If you want the primary source material, Schein's Organizational Culture and Leadership is the standard reference. It has gone through four editions and the later editions add more case material from non-American contexts, which the earlier versions lacked. There are also decent supplementary readings on subculture conflict and cultural measurement tools, but most of those tools have questionable reliability. The survey instruments that try to quantify culture tend to measure artifacts and espoused values reliably and underlying assumptions not at all, which means they give you data that tells you almost nothing useful about the actual operating system of the organization.